John L., a 42-year-old man, is in for his first visit. He indicates that he has no significant past medical history and is not taking any medications. You are relieved to hear this, because your new practice standards allow only 30 minutes for a new patient, and this particular morning you are already running 45 minutes late.

John now reveals that he has been HIV-positive for eight years. Two years ago, at a friend's urging, he found out that his CD4 count was 750. He knew this was "a good count"; he has not been checked since. Over the past year he has heard through the media about new treatments and possible "cures" for HIV disease, and he has come in for re-evaluation.

John feels fine and has no constitutional symptoms of HIV infection. His examination is normal. He agrees to have baseline laboratory tests and makes an appointment to return in a month's time. His lab results include a CD4 count of 560 and a viral load of 8,500 particles/mL.

John fails to keep his scheduled appointment. When he does return, some weeks later, you discuss his test results with him. You speak to him about potential therapies, citing both their advantages and their potential toxicities. You make a point of stressing the importance of adherence to therapy. Given John's missed appointment, the latter issue is a matter of particular concern.

After much discussion, John begins treatment with a combination of ZDV and 3TC. He opts for a twice-a-day regimen with proven efficacy. He understands that there might be some side effects of ZDV therapy. He decides not to add a protease inhibitor to his regimen at this time. You accept his decision.

Four weeks later, you repeat John's blood work. His HIV RNA has fallen to undetectable levels (< 500 particles/mL as measured by a second-generation bDNA test). He will return for a follow-up appointment in approximately two months.

The Clinical Questions:

• Would you have used a protease inhibitor as initial antiretroviral therapy?
• Would your treatment recommendations vary if John's HIV RNA had been 17,500 rather than 8,500?
• How often would you recheck the patient's viral load?
• If you had included a protease inhibitor in John's initial therapeutic regimen and his viral load dropped to undetectable levels, how long would you continue triple therapy?

The Experts' Recommendations:

Dr. Molly Cooke, professor of medicine, UCSF Medical School, and member of the editorial advisory board of HIV Newsline:

Would you have used a protease inhibitor as initial antiretroviral therapy?

I recognize that many authorities now recommend triple-drug therapy in all patients starting antiretroviral treatment, but I would not have used a protease inhibitor as part of John's initial regimen. As a rule, I do not use protease inhibitors in patients with relatively low viral burden and/or an uncertain commitment to therapy. Protease inhibitor therapy adds substantial toxicity to the antiretroviral regimen, and if I can accomplish the desired suppression of viral burden without adding a protease inhibitor, I feel justified in withholding the more powerful drugs until later.

Would your treatment recommendations vary if John's HIV RNA had been 17,500 rather than 8,500?

No. We don't yet know if two-drug regimens are less successful in suppressing viral RNA to undetectable levels in patients with a baseline count of 17,500 particles/mL than in those with a baseline count of 8,500 particles/mL, and in the absence of definitive data I have tended to reserve initial triple-drug therapy for patients with baseline viral RNA levels in excess of 50,000 particles/mL. In this case my decision is influenced by my sense that John is not yet thoroughly committed to therapy -- which I regard as an essential component of effective antiretroviral treatment with complex, demanding, multidrug regimens.

How often would you recheck the patient's viral load?

After an initial assessment of the efficacy of the patient's regimen, three to four weeks after its initiation, my practice is to check viral load every three to four months -- or when the patient's clinical status changes.

If you had included a protease inhibitor in John's initial therapeutic regimen and his viral load dropped to undetectable levels, how long would you continue triple therapy?

Assuming John tolerated the regimen well, I would continue triple therapy until I saw evidence that it wasn't working. In most cases, this conclusion would be based on rising viral load, although a marked change in the patient's clinical status could also prompt an adjustment. The other circumstance in which it is necessary to alter a patient's regimen is when there is an overriding need to use a drug that produces a deleterious drug-drug interaction with one of the patient's antiretroviral agents. Typically, these drug-drug interactions occur in patients being treated for MAC or multidrug-resistant TB.

Dr. Judith Aberg, AIDS Program, San Francisco General Hospital, and member of the editorial advisory board of HIV Newsline:

Would you have used a protease inhibitor as initial antiretroviral therapy?

We do not yet know what the best initial therapy is for an asymptomatic individual with CD4 counts above 500 cells/mm3. There is a lot of interest at present in developing studies that evaluate the benefits of "tight" control of viral load (no detectable virus by second-generation bDNA testing) versus "loose" control (viral load < 5,000 particles/mL), but these studies will take years to complete.

Natural history studies of HIV progression have established that the mean time from seroconversion to the development of AIDS is eight years. If John L. has in fact been HIV-positive for eight years, his CD4 count of 560 and his viral load of 8,500 suggest a favorable prognosis. In fact, an argument could be made, in light of these laboratory values, that John should not begin antiretroviral therapy until his viral load exceeds 10,000 particles/mL, his CD4 count drops below 500, or he develops symptomatic disease. The therapeutic options, in a case like this one, are: ddI monotherapy, combination therapy with two nucleosides, or the combination of two nucleosides plus a protease inhibitor. I agree that the combination of ZDV and 3TC is appropriate initial therapy in this patient, particularly given that the patient made an informed decision about the chosen regimen.

Like Dr. Cooke, I believe that all decisions about initial therapy should take into account both the potential side effects of the drugs under consideration and the patient's preferences with regard to dosing regimens. Therapeutic regimens that include a protease inhibitor require a highly structured daily dosing schedule, and this can be very daunting to some patients.

Would your treatment recommendations vary if John's HIV RNA had been 17,500 rather than 8,500?

I would make the decision to use a protease inhibitor based on the patient's willingness to undertake a three-drug regimen, not a difference of this size in viral load. Clearly, if John's viral burden was greater than 50,000 particles/mL my discussions with him would place more emphasis on the benefits of triple-drug combinations. In a patient with a baseline viral load of 17,500 and any degree of reluctance to undertake the rigors of protease inhibitor therapy, I would opt to start with two nucleosides. I would repeat the viral load measurement in three to four weeks, and if the patient still had detectable virus, I would recommend adding a protease inhibitor to the regimen.

How often would you recheck the patient's viral load?

In 1996 the International Panel for Antiretroviral Therapy for HIV Infection recommended that viral load be tested at baseline and then three to four weeks after initiating therapy (or making any changes in that therapy). Thereafter the patient's viral load should be measured every three to four months.

If you had included a protease inhibitor in John's initial therapeutic regimen and his viral load dropped to undetectable levels, how long would you continue triple therapy?

There is, at this point, no scientific answer to this question. I personally have never discontinued a patient's therapy simply because that patient achieved persistently undetectable levels of virus. This question has been posed to the AIDS Clinical Trials Group, and a trial is being developed. ACTG 343 proposes to start patients on triple therapy and then, after 12 months of treatment, to randomize patients to one of three maintenance arms: ZDV plus 3TC, indinavir alone, or a combination of all three drugs.

After six months of follow-up, those patients whose viral burden remains undetectably low will be randomly chosen to go off therapy. (Their viral levels will be closely monitored throughout this period.)

It will be years before ACTG 343 gives us a clinically relevant answer to the question posed here. Until that time I plan to continue the regimens my patients are on, making changes as they are needed to keep those patients' viral burdens as low as possible.

Is there any benefit to adding an additional agent if the patient already has an undetectable viral load? I do not know. I do know of one patient in our clinic who elected to add a protease inhibitor to his regimen after his viral burden dropped to undetectable levels on a combination of two nucleosides. There is no way of knowing if this decision will change the course of his disease, but the patient does report that he feels better, and he notes increases in energy and in activity levels.

It is important to remember that such anecdotal reports are subjective. They often reflect transient improvements, and those benefits, whether actual or perceived, may not be in the best long-term interest of the patient.

Susan Shea, assistant clinical professor, UCSF School of Nursing, and member of the editorial advisory board of AIDS Care:

Would you have used a protease inhibitor as initial antiretroviral therapy?

Like my colleagues, Drs. Cooke and Aberg, I would need to know John's philosophy of treatment before making specific recommendations. Dr. Makadon tells us that John did not want to include a protease inhibitor in his initial antiretroviral regimen. Why not? Was he concerned about potential side effects? Was he apprehensive about the responsibilities and restrictions that such a regimen imposes?

During my first meeting with a new patient I spend a good deal of time exploring how the patient has been dealing with his chronic infection. In medical circles we speak of "alternative therapies" -- which implies an either/or choice on the patient's part. What we should call these survival strategies is "concurrent therapies" or "complementary therapies." Whether they involve specific nutritional programs, regular exercise, acupuncture, meditation and prayer, or even foot reflexology and Chinese herbs, these complementary therapies can contribute significantly to the patient's sense of well-being, and it is important for me to elicit this sort of information from the patient -- so that I can better understand the patient's frame of reference and philosophy of treatment.

What we do know about John L. is that he is reluctant to begin an antiretroviral regimen that includes a protease inhibitor. He has responded well to therapy with two nucleoside analogs -- as many patients continue to do -- and his response confirms the wisdom of Dr. Makadon's decision. One therapeutic option that John might be asked to consider in the future is adding a non-nucleoside reverse transcriptase inhibitor, either nevirapine or delavirdine, to his drug regimen. The NNRTIs appear to be work well in combination with nucleoside analogs in treatment-naïve patients with high CD4 counts, achieving significant reductions in viral burden and decreasing the likelihood of resistance.

Would your treatment recommendations vary if John's HIV RNA had been 17,500 rather than 8,500?

The data from the Multicenter AIDS Cohort Study suggest that a viral load above 5,000 particles/mL greatly increases the relative risk of progression to AIDS. I therefore concur with Dr. Aberg, and I too would recommend the addition of a protease inhibitor if John's viral burden did not drop below 5,000 after a month of therapy.

How often would you recheck the patient's viral load?

My practice is to check viral load whenever our team is considering a change in therapy. If a patient is clinically stable on a particular regimen, has an undetectable viral load and good quality of life, I would assay HIV RNA every 8 to 12 weeks.

If you had included a protease inhibitor in John's initial therapeutic regimen and his viral load dropped to undetectable levels, how long would you continue triple therapy?

Once a patient begins triple therapy, he and his primary care provider are committed to triple therapy -- at least until studies show that therapy can be reduced without promoting resistance. What my own experience has taught me is that treatments that focus exclusively on medications -- and neglect nutrition, immune modulation, exercise, stress reduction, spiritual needs, and good communication between the patient and his care partners -- are much less effective than those that consider the whole of the patient's life.

Harvey J. Makadon, M.D., is Vice President and Medical Director ofAmbulatory Care and Community Health, Beth Israel Deaconess Medical Center, Boston, MA.